Dosing of Continuous Fentanyl Infusions in Obese Children: A Population Pharmacokinetic Analysis

Abstract

Differences in fentanyl pharmacokinetics (PK) between obese and nonobese adults have previously been reported; however, the impact of childhood obesity on fentanyl PK is relatively unknown. We developed a population pharmacokinetic (PopPK) model using opportunistically collected samples from a cohort of predominately obese children receiving fentanyl per the standard of care. Using a probability-based approach, we evaluated the ability of different continuous infusion strategies to provide steady-state concentrations (Css ) within an analgesic concentration range (1-3 ng/mL). Fifty-three samples from 32 children were used for PopPK model development. Median (range) age and body weight of study participants were 13 years (2-19 years) and 52 kg (16-164 kg), respectively. The majority of children (94%) were obese. A 2-compartment model allometrically scaled by total body weight provided an appropriate fit to the data. Estimated typical clearance was 32.5 L/h (scaled to 70 kg). A fixed dose rate infusion of 1 µg/kg/h was associated with probabilities between 49% and 58% for achieving Css within target; however, the risk of achieving Css > 3 ng/mL increased with increasing body weight (15% at 16 kg vs 43% at 164 kg). A proposed model-based infusion strategy maintained consistent probabilities across the examined weight range for achieving Css within (58%) and above (20%) target. Use of an allometric relationship between weight and clearance was appropriate for describing the PK of intravenous fentanyl in our cohort of predominately obese children. Our proposed model-derived continuous infusion strategy maximized the probability of achieving target Css in children of varying weights.

Keywords: dosing; fentanyl; obesity; pediatric; population pharmacokinetics.

© 2019, The American College of Clinical Pharmacology.

Figures

Figure 1.

Goodness-of-fit plots for the final PopPK model (A) observed vs. individual predicted concentrations; (B) observed vs. population predicted concentrations; (C) conditional weighted residuals vs. population predicted concentrations; and (D) conditional weighted residuals vs. body weight. In A and B, the line of identity is depicted by the solid line. In all subplots, dotted lines depict locally estimated scatterplot smoothing (LOESS) lines.

Figure 2.

Probability-based assessments for fixed dose rate infusions Probabilities associated with (A) 1; (B) 2; and (C) 3 mcg/kg/h infusions for achieving steady-state plasma fentanyl concentrations of 3 ng/mL.

Figure 3.

Probability-based assessment for a model-based dosing scheme (A) model derived infusion rates to achieve a target steady-state concentration of 1.7 ng/mL [infusion rate (mcg/kg/h) = 2.283/(WTkg0.25)]; (B) probabilities for achieving steady-state plasma fentanyl concentrations of <1, 1–3, and >3 ng/mL.

Figure 4.

Typical clearance estimates Comparison of typical clearance estimates (A: absolute; B: weight-normalized) from the current analysis to two other pediatric PopPK models., Displayed clearance relationships span the range of weights that were common between the subjects from the current and published analyses.